Despite a vast literature on the habituation of visual attention in infants, there has been very little attention paid to critical questions about the nature of this basic psychological phenomenon. In particular, there has been little research and almost no formal theory concerning the time course of habituation in individual infants. The overarching goal of this research project is to develop and test a statistical model of the habituation process. In doing so, we propose to address the following questions: what is the form of the habituation function? In particular, what function, or suitable family of functions best describes the habituation process in individual infants? How can habituation be rigorously assessed? How can the issue of individual differences in habituation functions be captured within a unified habituation model framework? Do individual infants show distinct "types" of habituation patterns, as has been widely suspected, and how can these be identified rigorously? How can one use information from habituation to improve inference about subsequent performance during the test phase? Our efforts to address these questions will involve developing new techniques for measuring habituation and testing these methods in real time by collecting habituation data from infants and applying the techniques to the resulting data. The development of a model-based approach to infant visual habituation will provide improved tools for equating levels of familiarity among infants, determining when habituation has occurred, rigorously classifying patterns of response in individual infants, and for using habituation data to assess post-habituation responses.